• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

New PC wont post in specific ram slot

Joined
May 27, 2012
Messages
347 (0.07/day)
Location
Malaysia
System Name Elmo
Processor Ultra 9 285K
Motherboard MSI Z890 A Wifi
Cooling Arctic Liquid Freezer III PRO
Memory 64GB DDR5 6000mts
Video Card(s) MSI Gaming Trio RTX 5070Ti
Storage 2TB Samsung 990 Pro,1TB Samsung 860 Evo, Samsung 970 Evo Plus 500GB, 8TB DC HC510
Display(s) LG 240hz 27" 2560
Case Antec Flux Pro
Power Supply Corsair RM850X
Software Windows 10 Pro
MSI z890 board

Already tested in all slots Only A1/A2 boots.
Checked for bent pins
Installed latest bios
Checked all power cables

I cant seem to boot in slot b1/b2

A1 B1 - No post
A1 only - Post
A1 A2 - Post
B1 - No post
B2 - No Post
B1 B2 - No Post

Dbug LED Error 55 - No memory detected when it doesnt post.

Is it faulty ram or motherboard? my gut feeling tells me its a motherboard issue. Would like some confirmation before i initiate a warranty claim
 

Attachments

  • _DSC0364.JPG
    _DSC0364.JPG
    2.7 MB · Views: 52
  • _DSC0366.JPG
    _DSC0366.JPG
    3.3 MB · Views: 53
Last edited:
What does the manual say, which slots should be populated first?
Does A1,A2 and B1,B2 work?

Is the RAM in jedec settings, otherwise test b1,b2 with jedec timings to check the slots. Also look at the vram compability for that board.
 
A1 A2 Works. B1 B2 does not work.

Ram is compatible. Motherboard says A1 B1 For 2 sticks. But A1 B1 wont Post. Anything with B populated wont post.
 
I’d make sure stick is truly in on the B slot, try it with jedec timings, but it is starting to look like a HW thing, my guess is same as yours on the MB. Make sure to reset cmos when trying A1/B1 with jedec.
 
I see at least 12 Z890 motherboards by MSI. Specifically, which board do you have?

I note for the Z890 Gaming Plus WIFI board, there have been 5 BIOS Firmware updates this year alone, with the latest just a couple weeks ago on July 9th. Depending on how long ago that board left the factory, you may be several versions behind.

To be sure, I am NOT a fan of flashing the BIOS just because a new version is out there. "If it ain't broke, don't fix it!" is my general philosophy when it comes to electronics. So I generally do NOT update the BIOS unless the update fixes a problem I am having, or it adds support to new hardware I want to install (newly released CPUs or RAM, as examples).

My standard exception to that philosophy is with new builds. When doing a new build, I typically make sure I have the latest BIOS version. This is especially true for motherboards that have recently hit the market within the last 10 - 12 months. That is because the first year is when most "bugs" (if any) are discovered and fixed.
 
I see at least 12 Z890 motherboards by MSI. Specifically, which board do you have?

I note for the Z890 Gaming Plus WIFI board, there have been 5 BIOS Firmware updates this year alone, with the latest just a couple weeks ago on July 9th. Depending on how long ago that board left the factory, you may be several versions behind.

To be sure, I am NOT a fan of flashing the BIOS just because a new version is out there. "If it ain't broke, don't fix it!" is my general philosophy when it comes to electronics. So I generally do NOT update the BIOS unless the update fixes a problem I am having, or it adds support to new hardware I want to install (newly released CPUs or RAM, as examples).

My standard exception to that philosophy is with new builds. When doing a new build, I typically make sure I have the latest BIOS version. This is especially true for motherboards that have recently hit the market within the last 10 - 12 months. That is because the first year is when most "bugs" (if any) are discovered and fixed.
I have the MSI Z890-A Wifi
I did not update the bios initially, i only updated it after i realized there was a problem. Even after the update its still the same.

I’d make sure stick is truly in on the B slot, try it with jedec timings, but it is starting to look like a HW thing, my guess is same as yours on the MB. Make sure to reset cmos when trying A1/B1 with jedec.
it was bone stock when i did the first boot so no difference unfortunately..

Would anyone know how to determine if it’s a faulty cpu or motherboard? Keeping in mind that it does post on dim a1 a2
 
Only factory test facility have the resources and test equipment to do that.

Our only options are try that CPU in a different, compatible motherboard, or try a different compatible CPU in that board.
 
Please reseat the processor and check your standoff screws to see if the board isn't warped. It honestly sounds like your second memory channel isn't connected properly. If the issue persists, RMA the motherboard.
 
I second checking the CPU and mounting. What @Dr. Dro references has been more common as of late.

Make sure the cooler is level and the motherboard isn’t warped.

Are you using a contact frame or the stock ILM??
 
I’m us
I second checking the CPU and mounting. What @Dr. Dro references has been more common as of late.

Make sure the cooler is level and the motherboard isn’t warped.

Are you using a contact frame or the stock ILM??
Im Using a contact frame arctic liquid freezer iii. My hands are already sore from reseating and checking for bent pins. I also checked the back of the cpu to see if any of the gold pads are missing all looks good. I may start exploring the idea of an rma soon. Board doesn’t look warped.. unfortunately have no friends with z890 boards to test with.
 
Put the original socket frame back on see if that solves the problem
 
Last edited:
Update! I decided to run another test and enable 200s in the bios.

Oddly enough did 2 things differently.
I installed the contact frame slightly looser together with the cooler.
I enabled 200s in bios.

And it booted in A1 B1!

I have run a prime95 test but not sure if my temps are in a normal range. Maybe someone who knows can clarify if its within the normal range

Idle : 42-50c
Full load : 80-102c
 
I would suspect the cooler was over tightened and affected the traces to the Ram slots.
 
Update - I dont think the system is stable, im not sure what to look for to determine what is the root cause. If anyone has experience in troubleshooting this your help would be appreciated.

Please see the attached file from aida64
 

Attachments

  • stabilitytest.png
    stabilitytest.png
    36.7 KB · Views: 31
not the first time that i hear and see this on arrow lake... A1/2 works, B1/2 doesn't. it's mounting pressure issue.
had the same problem with the arctic contact frame and i had to reinstall it twice before it worked.
 
I have narrowed it down to memory issues, i have also disabled XMP and the memory benchmark still fails
 
If AID64 stress test fails you have serious problem. That program isn't very hard on the memory at all.

Once again to eliminate it is a memory or motherboard issue, put the original frame back on.
 
If AID64 stress test fails you have serious problem. That program isn't very hard on the memory at all.

Once again to eliminate it is a memory or motherboard issue, put the original frame back on.
This would have been the next item, fortunately everything is stable and running well. I did unscrew the heatsink two turns, and reseated the memory. I think you were 100% bang on that it had to do with the contact frame/heatsink being too tight. Since when did CPUs become so sensitive... On the other hand i really despise arctic cooler for not having a proper manual with instructions on how tight it should be mounted, there is no mention on the tension or how many turns. It was really frustrating and im glad its over, i want to thank everyone for all the helpful suggestions.
 
Im Using a contact frame arctic liquid freezer iii.
Very sorry about your situation.

A big problem with this "product" is that in the case of a new build (with less experienced builders) where nothing is known yet about the state of the mobo, CPU and RAM, the mounting of the contact frame introduces an unwanted variable into the equation -> if something turns out to be wrong/faulty how do you know for sure that the mounting didn't affect anything?

Even if you acquire/borrow a cooler with standard mounting you have to put the ILM back very carefully in order to make sure that the pressure is in the normal range, not too high, not too low so that you can identify which of the three main components are faulty.
Or identify that indeed the contact frame was the culprit by being mounted improperly with regards to pressure.
A lot of hassle, back and forth, and for what?
It was really frustrating and im glad its over, i want to thank everyone for all the helpful suggestions.
Look, honestly if you still have a return window ditch that b*tch. You don't need such a "product" in your build.

If you absolutely want to use a contact frame (you don't really need one like with LGA1700), get a properly designed one like the Thermalright one, and whatever AIO you want that has standard mounting.
 
Very sorry about your situation.

A big problem with this "product" is that in the case of a new build (with less experienced builders) where nothing is known yet about the state of the mobo, CPU and RAM, the mounting of the contact frame introduces an unwanted variable into the equation -> if something turns out to be wrong/faulty how do you know for sure that the mounting didn't affect anything?

Even if you acquire/borrow a cooler with standard mounting you have to put the ILM back very carefully in order to make sure that the pressure is in the normal range, not too high, not too low so that you can identify which of the three main components are faulty.
Or identify that indeed the contact frame was the culprit by being mounted improperly with regards to pressure.
A lot of hassle, back and forth, and for what?

Look, honestly if you still have a return window ditch that b*tch. You don't need such a "product" in your build.

If you absolutely want to use a contact frame (you don't really need one like with LGA1700), get a properly designed one like the Thermalright one, and whatever AIO you want that has standard mounting.
I actually had noctua ship me the mounts for lga 1851 for my old nhd15.. but because i wanted to still use my old pc i decided to get an AIO. Ive never used an AIO before so this is my first experience with using contact frame as well. I almost sent my motherboard in for RMA today, im glad i decided to do more testing before doing that.
 
I actually had noctua ship me the mounts for lga 1851 for my old nhd15.. but because i wanted to still use my old pc i decided to get an AIO. Ive never used an AIO before so this is my first experience with using contact frame as well. I almost sent my motherboard in for RMA today, im glad i decided to do more testing before doing that.
Some hassle... that's the thing with a non-standard mounting, you can't swap coolers easily so that you know the mounting is out of the question.
More over the fact that it can lead to wrong verdicts (like suspecting and then returning the board) and a lot of trouble after shows that this design approach is the wrong one. It should be acknowledged, not swept under the rug.
Glad it got sorted, and that your components are functional! :)
 
Some hassle... that's the thing with a non-standard mounting, you can't swap coolers easily so that you know the mounting is out of the question.
More over the fact that it can lead to wrong verdicts (like suspecting and then returning the board) and a lot of trouble after shows that this design approach is the wrong one. It should be acknowledged, not swept under the rug.
Glad it got sorted, and that your components are functional! :)
Yes, after this debacle i googled to see if anyone else had similar issues with arctic contact frame. I couldn't find anyone talking about it.. but it really needs to be highlighted. Intel should also do a better job with their ILM and manufacturers who use contact frames should have information about these common/uncommon issues.
 
Intel should also do a better job with their ILM
Well they did change it for LGA1851, it's not perfect but at least it makes the contact frame optional instead of mandatory for most people (enthusiasts will always want the last drop).
Here and here (@16:48).
 
Yes, after this debacle i googled to see if anyone else had similar issues with arctic contact frame. I couldn't find anyone talking about it.. but it really needs to be highlighted. Intel should also do a better job with their ILM and manufacturers who use contact frames should have information about these common/uncommon issues.
Yeah, when the contact frames first hit the market there was so much coverage and hoopla about them that it seems like Arctic assumed everybody knew about them, wanted them, and knew what to be careful about when using them. This was a mistake. It's a great cooler, but that's after you get the mount right, which seems to effect a lot of people.

Also, as someone who can relate to not booting or instability after over-tightening a cooler on a CPU with a contact frame...I understand the pain and frustration lol. I was using a water block that has no "bottoming out" type of mounting bracket, so you just have to tighten by feel or buy a little torque wrench and well, I wanted good pressure so I cranked it down haha. Turns out gentle is better for those. Live and learn.
 
Back
Top